J. Phy8iol. (1978), 281, pp. 325-338 With 6 text-ftgure8 Printed in Great Britain
325
VASCULAR AND METABOLIC RESPONSES TO ADRENERGIC STIMULATION IN ISOLATED CANINE SUBCUTANEOUS ADIPOSE TISSUE AT NORMAL AND REDUCED TEMPERATURE
BY PAUL HJEMDAHL AND ALF SOLLEVI From the Department of Pharmacology, Karolinaka Institutet, 5-104 01 Stockholm, Sweden
(Received 31 December 1977) SUMMARY
1. The circulatory and metabolic effects of temperature reduction were studied in autoperfused canine subcutaneous adipose tissue in situ. 2. Cooling the adipose tissue sufficiently to reduce venous effluent temperature by 5-6 'C decreased blood flow from an average of 6-4-441 ml. min-. 100 g-1. 3. Vasoconstrictor responses to sympathetic nerve stimulation (4 Hz) and injected noradrenaline (5 n-mole) were potentiated by cooling while vasodilator components of the vascular responses, such as autoregulatory escape and post-stimulatory hyperaemia, were virtually abolished by this treatment. 4. Oxygen uptake was reduced by cooling without signs of tissue hypoxia. This reduced oxygen demand may partly cause the decrease in adipose tissue blood flow. 5. Cooling inhibited glycerol mobilization from the adipose tissue during sympathetic nerve stimulation. Post-stimulatory lipolysis was, however, not inhibited. In vitro studies with 'perifused' rat fat cells suggest that this may be due to impaired inactivation of the lipolytic process, rather than to changes in transmitter removal, following stimulation at low temperature. 6. Cooling inhibited the mobilization of fatty acids more than that of glycerol, suggesting increased re-esterification of fatty acids within the tissue at low temperature. 7. It is concluded that cooling increases the sensitivity to vasoconstrictor stimuli and that inhibition of metabolic vasodilator mechanisms play a role for this effect. The simultaneous inhibition of activating and inactivating mechanisms could explain the unchanged vascular and lipolytic responses to brief stimuli. Some possible implications of the present findings for the physiology of adipose tissue during cooling are discussed. INTRODUCTION
Adrenergic responses in canine subcutaneous adipose tissue in situ have been studied in some detail. Sympathetic nerve stimulation or the intra-arterial (I.A.) administration of noradrenaline induces vasoconstriction mediated by a-adrenoceptors (Ngai, Rosell & Wallenberg, 1966; Fredholm, Oberg & Rosell, 1970). After a-adrenoceptor blockade these stimuli induce vasodilatation mediated by f,adrenoceptors (Ngai et al. 1966; Belfrage & Rosell, 1976). Furthermore, nerve
P. HJEMDAHL AND A. SOLLE VI 326 stimulation and intravenously administered catecholamines enhance lipolysis, oxygen consumption and glycogenolysis secondary to stimulation of fl-adrenoceptors (Fredholm, 1970). There is some evidence that metabolism in the canine subcutaneous adipose tissue influences the circulation. Thus, inhibition of adipose tissue metabolism by fadrenoceptor blockade increases maximal vasoconstriction due to noradrenaline, decreases autoregulatory escape and decreases poststimulatory hyperaemia (Belfrage, 1978). However, /J-adrenoceptor blocking agents inhibit also vascular fl-adrenoceptors and the interpretation of these experiments is complicated. Recently it was found that acidosis, which inhibits lipolysis but not vasodilatation caused by ,adrenoceptor activation, also potentiated noradrenaline-induced vasoconstriction in the adipose tissue (Hjemdahl & Fredholm, 1976). These results were interpreted as evidence for a role of a 'metabolic' f3-adrenoceptor in the regulation of adipose tissue circulation. In order to further investigate the influence of metabolism on adipose tissue vascular responses we have utilized the well known effect of hypothermia to lower the metabolic rate (Bigelow, Lindsay, Harrison, Gordon & Greenwood, 1950). The temperature of subcutaneous tissue is decreased when the ambient temperature is lowered (Bruck & Henzel, 1953). Therefore the study of the effects of local cooling of the canine subcutaneous adipose tissue may be of physiological relevance. Subcutaneous adipose tissue may play a dual role when the ambient temperature is lowered. First, it may act as an insulator of the core of the body, a function which presupposes a low blood flow to prevent heat dissipation. Secondly, it stores energy in the form of fatty acids, the mobilization of which requires an adequate blood flow to the tissue. In the present study we therefore studied the influence of local cooling on subcutaneous adipose tissue circulation, metabolic rate and on lipid mobilization. A preliminary account of some of the present findings has been reported elsewhere (Sollevi, Hjemdahl & Fredholm, 1975). METHODS
In vivo techniques The experiments were performed on sixteen female mongrel dogs, weighing 10-26 kg. The dogs were anaesthetized with sodium pentobarbitone (30 mg/kg i.v.) with supplements as required. Tracheotomy was performed and the dogs were mechanically ventilated with a Braun Melsungen model 74052 respirator. Subcutaneous adipose tissue in the inguinal region was subsequently isolated from skin and other surrounding tissues as described by Rosell (1966). This provided an adipose tissue preparation connected to the animal by one artery, one vein and one nerve containing adrenergic fibres. The weight of the preparation was between 30 and 190 g (average 52 g). The adipose tissue was enclosed in a saline-perfused chamber the temperature of which was maintained at either 370C or 26-270C as measured by a thermistor (Ellab Instruments, Copenhagen). Arterial blood was diverted from the femoral artery to the adipose tissue via a drop counter and venous outflow was returned to the femoral vein via a polyethylene catheter containing a three-way stopcock, which was used for venous blood sampling. A thermistor was also inserted into the catheter for venous blood temperature monitoring. Cooling the saline-perfused plethysmograph from 37 to 27 reduced the venous effluent temperature from approximately 33 0C to 27-280C. This cooling procedure thus reduced the temperature of the effluent blood by 5-6 0C. Heparin (2500 iLu./kg) was administered before the cannulation procedure, i.e. at least 1 hr before the first experimental run. Fluid losses due to sampling and trauma were replaced with isotonic saline. Systemic blood pressure was measured with Statham P23AC transducers and
°C
COOLING IN SUBCUTANEOUS TISSUE
327
recorded together with adipose tissue blood flow on a Grass model 7B polygraph. Vascular conductance was calculated by dividing adipose tissue blood flow by the blood pressure. The nerve supplying the adipose tissue was sectioned in all experiments. When stimulation was performed, the distal part of the sectioned nerve was placed on a bipolar silver electrode and protected from drying with Plastibase (Squibb). The nerve was stimulated at 4 Hz during periods of 5 or 30 min with impulses of supramaximal duration (1 msec) and intensity (12-14 V), which were delivered by a Grass model S4 stimulator. In four experiments 5 x 10-9 mole noradrenaline (L-norepinephrine hydrochloride, Sigma) was administered to the adipose tissue by close intraarterial injection. In two experiments involving nerve stimulation both inguinal fat pads were isolated. One of the adipose tissue preparations was pretreated with cocaine 200-400 berg i.A. before each observation period, while the contralateral preparation served as control. Samples of arterial blood and venous blood from the adipose tissue were collected into ice-cold plastic tubes. Aliquots of whole blood were immediately removed for the determination of lactate and pyruvate (TCC, TCB, Boehringer & Sohn, Mannheim). The samples were then centrifuged and plasma was removed for the determination of glycerol (Laurell & Tibbling, 1966), FFA (Laurell & Tibbling, 1967) and glucose (with commercially available glucose oxidase reagent, Glox, KABI, Stockholm). The venous sampling followed a standardized pattern, as illustrated in Figs. 1 and 3. Arterial samples were withdrawn at regular intervals for the determination of the hematocrit and the above-mentioned metabolites. The uptake or release of the various metabolites was subsequently calculated on the basis of arterio-venous concentration differences and blood or plasma flow values per 100 g tissue. In experiments where the oxygen uptake of the adipose tissue was determined, blood was collected under paraffin oil. The haemoglobin content was determined by a commercially available procedure (Aculute, Ortho diagnostics). Arterial and venous pH, pco, and po, were determined by a Radiometer (Copenhagen) BMS 3 MK2 blood gas analyser. The oxygen content of the blood samples was subsequently obtained from a nomogram.
In vitro techniques Isolated fat cells were prepared essentially according to Rodbell (1964). Epididymal fat pads from three to five male rats (Sprague Dawley, Anticimex strain, 180-200 g) were pooled in each experiment. After mincing, the tissue was incubated in Krebs-Ringer phosphate buffer pH 7.4, containing half the recommended calcium concentration, 5.5 mM-glucose, bovine serum albumin 30 mg/ml. (Fraction V, Sigma, St Louis) and crude bacterial collagenase 3 mg/ml. (Lot 44B231, Worthington Biochem., Freehold) for 40-60 min. After filtering through gauze and washing twice in the same medium without collagenase the cells were used either for incubation or perifusion experiments. In the incubation experiments 0 5 ml. of a concentrated cell suspension were added to 2 ml. buffer in plastic vials to give a final cell concentration of 40000 cells/ml. The cells were incubated at 37 or 27 0C in shaking water-baths. The cells were preincubated for 10 min before the addition of noradrenaline. At different time intervals after the addition of noradrenaline, aliquots of cells and medium were taken for the determination of glycerol (Laurell & Tibbling, 1966) and cyclic AMP (Brown, Ekins & Albano, 1972). The treatment of samples for the cyclic AMP determination has been described elsewhere (Fredholm & Hjemdahl, 1976). For the perifusion experiments 1 ml. packed fat cells were added to 2-2-5 ml. Krebs-Ringer phosphate buffer pH 7-4 in a temperature-controlled plastic chamber essentially according to Allen, Largis, Miller & Ashmore (1973). The floating fat cells in the chamber were subsequently perifused with the same medium containing albumin 10 mg/ml. at a rate of 2 ml./min. The chamber and the perifusion medium were kept at 37 or 27 'C. The perifusate was collected continuously and subsequently used for the determination of glycerol. In order to increase the sensitivity of the glycerol assay the deproteinization step was excluded. Standards and blanks were therefore prepared in the same medium and yielded expected values. Noradrenaline was infused via a side-arm to give a final concentration of 10 FM in the chamber. Conventional statistical methods were used to calculate means (i), standard deviations (S.) and standard errors of the mean (Si). Hypotheses were tested by Student's t test for paired or unpaired variates or by the Wilcoxon test in case the variates were not normally distributed.
P. HJEMDAHL AND A. SOLLE VI
328
RESULTS
Va8cular responses The resting blood flow in the denervated adipose tissue averaged 6-4 ml. min-'. 100 g-1 (range 3.0-13.0) during control conditions. When the temperature was reduced blood flow decreased to 4-1 ml. min-. 100 g-1 (range 2.0-7.7), which corresponded to a decrease in vascular conductance of 33-2 + 5-5 % (n = 16, P < 0-001). TABLE 1. Vascular responses to noradrenaline injections (5 n-mole) in four experiments and to sympathetic nerve stimulation (5 min, 4 Hz) in six experiments. Peak vasoconstriction and peak hyperaemia are expressed as % of prestimulatory vascular conductance. Mean values + s.E. are given Noradrenaline Nerve stimulation Peak vasoconstriction
(P Peak hyperaemia
27 0C 21-8 + 4-3
37 0C 25-0 ± 3-8
325
VASCULAR AND METABOLIC RESPONSES TO ADRENERGIC STIMULATION IN ISOLATED CANINE SUBCUTANEOUS ADIPOSE TISSUE AT NORMAL AND REDUCED TEMPERATURE
BY PAUL HJEMDAHL AND ALF SOLLEVI From the Department of Pharmacology, Karolinaka Institutet, 5-104 01 Stockholm, Sweden
(Received 31 December 1977) SUMMARY
1. The circulatory and metabolic effects of temperature reduction were studied in autoperfused canine subcutaneous adipose tissue in situ. 2. Cooling the adipose tissue sufficiently to reduce venous effluent temperature by 5-6 'C decreased blood flow from an average of 6-4-441 ml. min-. 100 g-1. 3. Vasoconstrictor responses to sympathetic nerve stimulation (4 Hz) and injected noradrenaline (5 n-mole) were potentiated by cooling while vasodilator components of the vascular responses, such as autoregulatory escape and post-stimulatory hyperaemia, were virtually abolished by this treatment. 4. Oxygen uptake was reduced by cooling without signs of tissue hypoxia. This reduced oxygen demand may partly cause the decrease in adipose tissue blood flow. 5. Cooling inhibited glycerol mobilization from the adipose tissue during sympathetic nerve stimulation. Post-stimulatory lipolysis was, however, not inhibited. In vitro studies with 'perifused' rat fat cells suggest that this may be due to impaired inactivation of the lipolytic process, rather than to changes in transmitter removal, following stimulation at low temperature. 6. Cooling inhibited the mobilization of fatty acids more than that of glycerol, suggesting increased re-esterification of fatty acids within the tissue at low temperature. 7. It is concluded that cooling increases the sensitivity to vasoconstrictor stimuli and that inhibition of metabolic vasodilator mechanisms play a role for this effect. The simultaneous inhibition of activating and inactivating mechanisms could explain the unchanged vascular and lipolytic responses to brief stimuli. Some possible implications of the present findings for the physiology of adipose tissue during cooling are discussed. INTRODUCTION
Adrenergic responses in canine subcutaneous adipose tissue in situ have been studied in some detail. Sympathetic nerve stimulation or the intra-arterial (I.A.) administration of noradrenaline induces vasoconstriction mediated by a-adrenoceptors (Ngai, Rosell & Wallenberg, 1966; Fredholm, Oberg & Rosell, 1970). After a-adrenoceptor blockade these stimuli induce vasodilatation mediated by f,adrenoceptors (Ngai et al. 1966; Belfrage & Rosell, 1976). Furthermore, nerve
P. HJEMDAHL AND A. SOLLE VI 326 stimulation and intravenously administered catecholamines enhance lipolysis, oxygen consumption and glycogenolysis secondary to stimulation of fl-adrenoceptors (Fredholm, 1970). There is some evidence that metabolism in the canine subcutaneous adipose tissue influences the circulation. Thus, inhibition of adipose tissue metabolism by fadrenoceptor blockade increases maximal vasoconstriction due to noradrenaline, decreases autoregulatory escape and decreases poststimulatory hyperaemia (Belfrage, 1978). However, /J-adrenoceptor blocking agents inhibit also vascular fl-adrenoceptors and the interpretation of these experiments is complicated. Recently it was found that acidosis, which inhibits lipolysis but not vasodilatation caused by ,adrenoceptor activation, also potentiated noradrenaline-induced vasoconstriction in the adipose tissue (Hjemdahl & Fredholm, 1976). These results were interpreted as evidence for a role of a 'metabolic' f3-adrenoceptor in the regulation of adipose tissue circulation. In order to further investigate the influence of metabolism on adipose tissue vascular responses we have utilized the well known effect of hypothermia to lower the metabolic rate (Bigelow, Lindsay, Harrison, Gordon & Greenwood, 1950). The temperature of subcutaneous tissue is decreased when the ambient temperature is lowered (Bruck & Henzel, 1953). Therefore the study of the effects of local cooling of the canine subcutaneous adipose tissue may be of physiological relevance. Subcutaneous adipose tissue may play a dual role when the ambient temperature is lowered. First, it may act as an insulator of the core of the body, a function which presupposes a low blood flow to prevent heat dissipation. Secondly, it stores energy in the form of fatty acids, the mobilization of which requires an adequate blood flow to the tissue. In the present study we therefore studied the influence of local cooling on subcutaneous adipose tissue circulation, metabolic rate and on lipid mobilization. A preliminary account of some of the present findings has been reported elsewhere (Sollevi, Hjemdahl & Fredholm, 1975). METHODS
In vivo techniques The experiments were performed on sixteen female mongrel dogs, weighing 10-26 kg. The dogs were anaesthetized with sodium pentobarbitone (30 mg/kg i.v.) with supplements as required. Tracheotomy was performed and the dogs were mechanically ventilated with a Braun Melsungen model 74052 respirator. Subcutaneous adipose tissue in the inguinal region was subsequently isolated from skin and other surrounding tissues as described by Rosell (1966). This provided an adipose tissue preparation connected to the animal by one artery, one vein and one nerve containing adrenergic fibres. The weight of the preparation was between 30 and 190 g (average 52 g). The adipose tissue was enclosed in a saline-perfused chamber the temperature of which was maintained at either 370C or 26-270C as measured by a thermistor (Ellab Instruments, Copenhagen). Arterial blood was diverted from the femoral artery to the adipose tissue via a drop counter and venous outflow was returned to the femoral vein via a polyethylene catheter containing a three-way stopcock, which was used for venous blood sampling. A thermistor was also inserted into the catheter for venous blood temperature monitoring. Cooling the saline-perfused plethysmograph from 37 to 27 reduced the venous effluent temperature from approximately 33 0C to 27-280C. This cooling procedure thus reduced the temperature of the effluent blood by 5-6 0C. Heparin (2500 iLu./kg) was administered before the cannulation procedure, i.e. at least 1 hr before the first experimental run. Fluid losses due to sampling and trauma were replaced with isotonic saline. Systemic blood pressure was measured with Statham P23AC transducers and
°C
COOLING IN SUBCUTANEOUS TISSUE
327
recorded together with adipose tissue blood flow on a Grass model 7B polygraph. Vascular conductance was calculated by dividing adipose tissue blood flow by the blood pressure. The nerve supplying the adipose tissue was sectioned in all experiments. When stimulation was performed, the distal part of the sectioned nerve was placed on a bipolar silver electrode and protected from drying with Plastibase (Squibb). The nerve was stimulated at 4 Hz during periods of 5 or 30 min with impulses of supramaximal duration (1 msec) and intensity (12-14 V), which were delivered by a Grass model S4 stimulator. In four experiments 5 x 10-9 mole noradrenaline (L-norepinephrine hydrochloride, Sigma) was administered to the adipose tissue by close intraarterial injection. In two experiments involving nerve stimulation both inguinal fat pads were isolated. One of the adipose tissue preparations was pretreated with cocaine 200-400 berg i.A. before each observation period, while the contralateral preparation served as control. Samples of arterial blood and venous blood from the adipose tissue were collected into ice-cold plastic tubes. Aliquots of whole blood were immediately removed for the determination of lactate and pyruvate (TCC, TCB, Boehringer & Sohn, Mannheim). The samples were then centrifuged and plasma was removed for the determination of glycerol (Laurell & Tibbling, 1966), FFA (Laurell & Tibbling, 1967) and glucose (with commercially available glucose oxidase reagent, Glox, KABI, Stockholm). The venous sampling followed a standardized pattern, as illustrated in Figs. 1 and 3. Arterial samples were withdrawn at regular intervals for the determination of the hematocrit and the above-mentioned metabolites. The uptake or release of the various metabolites was subsequently calculated on the basis of arterio-venous concentration differences and blood or plasma flow values per 100 g tissue. In experiments where the oxygen uptake of the adipose tissue was determined, blood was collected under paraffin oil. The haemoglobin content was determined by a commercially available procedure (Aculute, Ortho diagnostics). Arterial and venous pH, pco, and po, were determined by a Radiometer (Copenhagen) BMS 3 MK2 blood gas analyser. The oxygen content of the blood samples was subsequently obtained from a nomogram.
In vitro techniques Isolated fat cells were prepared essentially according to Rodbell (1964). Epididymal fat pads from three to five male rats (Sprague Dawley, Anticimex strain, 180-200 g) were pooled in each experiment. After mincing, the tissue was incubated in Krebs-Ringer phosphate buffer pH 7.4, containing half the recommended calcium concentration, 5.5 mM-glucose, bovine serum albumin 30 mg/ml. (Fraction V, Sigma, St Louis) and crude bacterial collagenase 3 mg/ml. (Lot 44B231, Worthington Biochem., Freehold) for 40-60 min. After filtering through gauze and washing twice in the same medium without collagenase the cells were used either for incubation or perifusion experiments. In the incubation experiments 0 5 ml. of a concentrated cell suspension were added to 2 ml. buffer in plastic vials to give a final cell concentration of 40000 cells/ml. The cells were incubated at 37 or 27 0C in shaking water-baths. The cells were preincubated for 10 min before the addition of noradrenaline. At different time intervals after the addition of noradrenaline, aliquots of cells and medium were taken for the determination of glycerol (Laurell & Tibbling, 1966) and cyclic AMP (Brown, Ekins & Albano, 1972). The treatment of samples for the cyclic AMP determination has been described elsewhere (Fredholm & Hjemdahl, 1976). For the perifusion experiments 1 ml. packed fat cells were added to 2-2-5 ml. Krebs-Ringer phosphate buffer pH 7-4 in a temperature-controlled plastic chamber essentially according to Allen, Largis, Miller & Ashmore (1973). The floating fat cells in the chamber were subsequently perifused with the same medium containing albumin 10 mg/ml. at a rate of 2 ml./min. The chamber and the perifusion medium were kept at 37 or 27 'C. The perifusate was collected continuously and subsequently used for the determination of glycerol. In order to increase the sensitivity of the glycerol assay the deproteinization step was excluded. Standards and blanks were therefore prepared in the same medium and yielded expected values. Noradrenaline was infused via a side-arm to give a final concentration of 10 FM in the chamber. Conventional statistical methods were used to calculate means (i), standard deviations (S.) and standard errors of the mean (Si). Hypotheses were tested by Student's t test for paired or unpaired variates or by the Wilcoxon test in case the variates were not normally distributed.
P. HJEMDAHL AND A. SOLLE VI
328
RESULTS
Va8cular responses The resting blood flow in the denervated adipose tissue averaged 6-4 ml. min-'. 100 g-1 (range 3.0-13.0) during control conditions. When the temperature was reduced blood flow decreased to 4-1 ml. min-. 100 g-1 (range 2.0-7.7), which corresponded to a decrease in vascular conductance of 33-2 + 5-5 % (n = 16, P < 0-001). TABLE 1. Vascular responses to noradrenaline injections (5 n-mole) in four experiments and to sympathetic nerve stimulation (5 min, 4 Hz) in six experiments. Peak vasoconstriction and peak hyperaemia are expressed as % of prestimulatory vascular conductance. Mean values + s.E. are given Noradrenaline Nerve stimulation Peak vasoconstriction
(P Peak hyperaemia
27 0C 21-8 + 4-3
37 0C 25-0 ± 3-8
